JPS6184515A - Method and device for obtaining distance up to plurality of point on object - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for determining the distance of an object used in fields such as robot vision, and an apparatus used therefor.
Particularly useful for creating dense range maps at high speed without the use of any moving mechanisms, the present invention includes a projector that projects a continuous striped pattern of light and a TV camera that captures the striped scene. is used. Video frames are decoded in two's complement domain to obtain a range map.

Vision plays an important role in automating a wide variety of industrial and military tasks. Three-dimensional vision is an essential sensory ability for closing feedback loops, adapting systems to situations, and working in unstructured environments. If the system controller wants to obtain more reliable information, not only the well-known 2D image forming ability but also the 3D ability will be very useful in improving productivity and boosting factors. There are many application fields, such as 3D component inspection, inventory investigation, 3D operation of autonomous vehicles, and monitoring of 1M slave power plants. Many efforts have been made to obtain distance information mainly in the following two technical areas.

The first area is the so-called r2. X-DJ, a pseudo three-dimensional 3D technique, which avoids distance # measurements of all resolvable pixels in the field of view. 1 instead
A normal grayscale two-dimensional image is combined with a priori knowledge with the help of an artificial brain to estimate the range profile. Alternatively, a two-dimensional image is combined with a rough sense of distance [8 constants] ranging from an average distance # measurement of one point to measurement of several selected points. The fraction X indicates how faithfully a particular technique approaches a three-dimensional system. The strength of this field of technology is the speed with which data can be obtained. The disadvantage is that blurring is inevitable when used in an unstructured environment.

The second field of distance measurement is so to speak r3. o-DJ, i.e. pure, is a three-dimensional technique, where distance measurements are performed for all pixels in the field of view. According to this technique, a "dense range map" can be obtained mainly by using active systematic light irradiation. Since this technique is a metrological measurement, it is independent of the blurring caused by estimates that is inevitable with the r2-X-DJ method. This technical field can be further divided into three technologies. The first technique is
A single light beam scanning technique uses mechanical beam steering equipment to make point-by-point distance # measurements by estimating the time-of-flight, or phase, of the encoded return beam. This technique is suitable for distant objects. This is because the given light is most effectively used at a specific point within the field of view to maximize the S/N ratio. A disadvantage of this technique is slow speed due to the time required for scanning and the need to use mechanical elements for scanning.

Distance #Measurement Method "3.0 ~ The second technique in the DJ field is to project a stripe instead of a point onto the target space at an oblique angle θ.
A TV camera in the Z meeting captures the subject with a two-dimensional detection array. Since the straight stripes appear deformed in the video frame due to the oblique angle of the projection, this deformation is converted into a surface profile using post-data acquisition methods or simple trigonometry. This technique has a particular advantage over the last one above, namely that it eliminates both the undesirable mechanical scanner and complex phase measurements. This technology also

It is particularly suited for applications such as inspection of parts being conveyed on a belt conveyor or as a seam follower in optical welding. However, this technology also has the drawback of low speed.

The entire frame time is required to estimate the result of one stripe. If you try to cover the scene with 256 tree stripes, it will take 1 normal frame time 39
Assuming m5ec, it will take at least 7 hours just to record the data.
It takes seconds. As such, this technique is often too slow for real-time data collection applications.

r3. o - The third technique of the DJ distance measurement method is.

Using multi-stripe pattern projection, one video frame can be used to obtain the entire stripe deformation since one pattern covers the entire field of view. Therefore.

It is a highly efficient data collection technology. However, this is only true when the deformed stripes are continuous functions. If there are discontinuities in the surface profile, the stripes will not have continuous boundaries and will be very indistinct. This technique is thus limited to the analysis of relatively smooth surfaces and the evaluation of local slopes. This is not an ideal method for reliable object recognition in an unstructured environment.

Margin below The question here is how to resolve the dilemma raised by the second and third techniques in the r 3.0-D J field of distance measurement methods. That is, if the measurement speed is improved, it will become blurred, and if it is made clearer, the speed will decrease.

The main object of the present invention is to provide a method and apparatus for creating dense range maps at high speed without mechanical movement or blurring.

The above-mentioned objectives and other objectives can be achieved by
This is accomplished by projecting a continuous pattern of stripes of light onto an object at an oblique angle to the V-right camera optical axis, and the camera capturing the projected light reflected from the object. The light stripes are binary encoded by selectively brightening and darkening each stripe in a unique sequence of continuous patterns. The entire target space is covered with a pattern of continuous stripes, and the number of stripes is selected to achieve the desired resolution.The number of stripes required to uniquely encode each stripe is set with 2 of the number of stripes as a constant. is equal to the logarithm of

The reflected light captured and uniquely encoded by each pixel of the TV right camera is indicative of the particular stripe illuminating the object, and thus the location of the object. Object x-
The Y coordinate can be easily obtained from the coordinates of both systems receiving the reflected light signal. The Z-axis coordinate is determined from the X-axis coordinate of the pixel and the known position of the light illumination stripe.

Although conventional trigonometry can be used to determine the distance of objects such as TV right cameras from pixel coordinate information and known tracks of encoded light stripes, the present invention provides a particularly simple method for determining this distance. Provides a fast configuration. A three-dimensional Cartesian coordinate system is created in the object space, and its X and Y axes are in equilibrium with the X-Y plane of the TV right camera, and the X axis is parallel to the optical axis of the camera. A stripe of light is projected through this coordinate system, with the stripe parallel to the Y axis and intersecting the X axis and the X axis at an oblique angle. The X-axis and the X-axis are graduated by the diagonal width of the stripe with respect to each axis. Furthermore, the camera pixel array
The axis scale is determined by the X-axis scale of the coordinate system of the target space.

Using this coordinate system, the stripes are numbered and encoded sequentially from the origin. In this way, for example, the D-th stripe intersects both axes in units of D from the origin on the scales of the X-axis and the X-axis. Since the D-th stripe is encoded in binary notation, the distance from the origin of the object space coordinate system of the object it illuminates is determined by the received X coordinate of the reflected D-th stripe light. It can be easily obtained by subtracting from the encoded signal.

In a preferred embodiment, the origin of the object space coordinates is located near the center of the object space, and the stripe coordinates are used to project a zero-order stripe pattern through the origin, collect data, and calculate the range: In order to be able to use a typical digital computer,
Use two's complement representation.

If a shallow field of view is required compared to the X dimension of the object space, the stripe pattern may be repeated as many times as possible in parallel. It also includes. According to this concept, reliability is improved by removing error data caused by unexpected movement of the object or secondary reflections. By using an odd parity scheme, there is no blind spot that would occur if there were a 0th order stripe through the origin.

The present invention, both as a method and apparatus for implementing the techniques described above, provides a number of particular advantages, including high reliability, high speed, digital compatibility, gray scale information and high range resolution. These advantages will be fully explained in the following description of the preferred embodiment.

Support-1-1 FIG. 1 illustrates the arrangement of equipment for carrying out the invention. This device consists of a projector 1 that projects a continuous pattern of light stripes onto a space 3 containing an object (the position of which will be determined later), and for each pattern of light stripes, light reflected from an object within the object space. A conventional TV right camera records the level of the TV. This camera 5 typically includes several hundred detectors arranged in an X-array. A pixel is each discrete point at which one of these detectors can measure the intensity of light. A general-purpose digital computer 7 controls the creation of a light encoding pattern by the projector 1, and changes the shade of the light intensity detected by each pixel of the TV left camera into a value for the intensity below a given threshold value. When the intensity exceeds 0 or a threshold value, it changes to a binary signal with a value of 1. The binary signal generated at each pixel for each pattern of the light stripe is stored, and this stored signal is further processed. to find the distance of the object from the TV left camera.

As shown in FIG. 2, the projector 1 is arranged geometrically with respect to the camera 5 such that its projection axis 8 is aligned with the optical axis 11 of the camera 5 due to the scattering of the surface. Therefore, the light from the projector 1 that illuminates the object 13.15 is
The projector 1 reflected at θ and detected by the TV left camera creates a pattern of light stripes 17, as illustrated in FIG. The stripes 17 are formed in the direction perpendicular to the plane defined by the axis 8.11.
Each stripe 17 is a sequence of patterns generated by the projector l, with illuminated areas being bright and non-illuminated areas being dark. Each stripe 17 is encoded into a unique binary number representing the state for each pattern in the sequence: ``1'' for bright areas and rOJ for dark areas. - By way of example, FIG. 3 illustrates the division of the projection pattern into 18 consecutive but scattered vertical stripes of equal width. In a typical system, the number of stripes would be much higher than those illustrated, but in any case they would be chosen to give the desired resolution, determined by the width of the stripes, as will be clear from the description below. will be done. Each of the stripes 17 is assigned a number starting at -8 on the left and increasing by one to the neighbor up to +7. Using two's complement notation, -8 is 10θ
O, -7 is 1001, and 0111 for 7
is expressed up to. In the above arrangement, the binary numbers assigned to each stripe, which would require 4 bits for each number to uniquely identify a 16-tree stripe, are shown below the stripe in Figure 3. Yes, from the most significant bit (MSB) to the least significant bit (LSB) vertically downward.
).

To create the four bits required to encode each stripe, projector l creates four different stripe patterns as illustrated in FIGS. 4-7. The highest numbered bee 2 representing 18 stripes
In the pattern of Figure 1 shown in Figure 4, which forms stripes according to the MSB, the stripes on the right are all dark (with the value ``0'') and the stripes on the left are all light (with the value rlJ). ), as a result, 2
Composite stripes of wood (one light and the other dark) are essentially formed. In the second pattern illustrated in FIG. 5, the spatial frequency of the four composite stripes is twice that of that in FIG. 4, but only half the width; similarly illustrated in FIG. In the third pattern, the spatial frequency of the composite stripe is four times that of the most significant bit, and the width of the composite stripe itself is 1/4. In the fourth pattern illustrated in FIG. 7, representing the least significant bit, adjacent stripes are in the opposite state and the spatial frequency is eight times that of the most significant bit.

The camera captures the reflections from the object 13.15 in which the four striped patterns of FIGS. 4 to 7 occur in binary format (see FIG. 3). By knowing the history of reflected light received by each pixel of the TV right camera, the stripe number of each point can be determined. In other words, the binary value obtained at each pixel represents which stripe is illuminating the object that is reflecting light to that pixel. It should be noted here that no correlation is required between the number of TV right camera pixels and the number of stripes, and that the stripes are at least as wide as the pixels.

The orientation of the projector 1 is adjusted so that it faces the center of the target space, and its axis intersects the camera light @ll at an oblique angle θ at the origin of the x-Z coordinates (see FIG. 2). Figure 8 shows an enlarged view of the area around object 13.15 in Figure 2, superimposing the x-Z coordinates and projecting the stripe pattern of the least significant bit in Figure 7 onto the drawing. It has been done.

As can be seen from Figure 8, the stripe number 0 is always x-
It passes through the origin of the Z plane. In this way, if there is a light reflecting plane at the origin.

Dark - dark, dark, dark, and exposure occurs. If the reflective surface moves one unit away from the origin along the Z axis (i.e., the distance increases), it is illuminated by the first stripe, creating a unique sequence of dark-dark-dark-light and exposure, creating a +1 stripe. happens. From the above it can be quickly deduced that there is a linear relationship between stripe order and reflective surface. The larger the stripe number, the greater the distance between the scattering surfaces. In this way, for a scattering surface along the Z-axis, the distance can be read directly on the digital display as a binary number created by a stripe intersecting the Z-axis at the distance of the reflective surface.

The inventors have so far devised a general and simple decoding method that allows determining the distance of a reflective surface at every point in the object space. Assume that there is an arbitrary scattering point P in the target space xyz. The function of the camera is to move point P to X-Y.
It is to project to the coordinates. Therefore, the pixel illuminated by the light reflected from the scattering surface at point P is digitally
It is easy to determine the X and Y values by simply determining the horizontal and vertical addresses within the array. The question here is only how to obtain the Z value using address information and history data of reflected light. To answer this question, a simple geometrical analysis can be performed in the X-Z plane containing point P, as illustrated in FIG. From the history of the reflected light for the camera pixel aligned with point P, it can be seen that the reflected light has a stripe number.

It can be seen that point P exists somewhere in the ray trace 13 of the D-th stripe.

It can be seen from FIG. 9 that the following equation holds true for straight line 18, ie, trace No. 0.

Z=AZ 10 Here, A is the slope of the straight line 19, which is negative in this case, and is the X-intercept. Relative scales are chosen for the X and X axes so that each stripe trace intersects each axis by the number of units corresponding to the stripe number, so the intersection of line 18 with the X axis is also at a distance from the origin. It is in. (Referring to FIG. 8, stripe number 5 intersects the X axis at X=5 and the X axis at 2-5.) Therefore, the slope A of straight line 19 is equal to -1. Substituting this into equation (1), the general equation for the X coordinate of an arbitrary point along the D-th stripe 18 is as follows.

z -n -x Regarding point P, since the X coordinate of the pixel that receives the reflected light from this point P is known, the relative distance (from the origin of the coordinate system) can be calculated from the encoded signal received at that pixel. It is immediately obtained by subtracting the X value.

When this technique is applied to an actual problem, the relative distance between points A and B of object 13.15 in Figure 8 can be found at 0 point A.
The X coordinate obtained from the X coordinate of the pixel adjusted to is +5, that is, 0101 in binary notation. As shown in Figure 8.

Point A is illuminated with +7 stripes. This number is
It is determined from the exposure experience of point A, which is dark-bright-one-bright, that is, 0111 in binary notation. These numbers are (2
), at point A: D = 0111 X = 0101 - Z = 0010 = 2 (decimal notation) In other words, the X coordinate is 2 in the basic lO base system. In this way, the chain point The relative distance of can be easily and quickly determined.

To explain that the technique also applies to negative numbers, consider point B whose Y coordinate is -2, that is, 1110 in two's complement notation. This point B is illuminated with 16 stripes coded as bright-dark-one-light-one-dark and expressed in binary 1010. Substituting these values into equation (2), at point B: D = 1010 X = 1110- 11000 = -4 (in decimal notation) This means that the It shows that there is.

If the target space to be covered is relatively shallow in the Z direction and wide in the X direction, the chosen sequence of stripe patterns can be repeated in the X direction as many times as necessary. For example, if 16 units in the Z direction are sufficient to cover the required depth of field, but 32 units are required in the X direction, then the 16 stripe pattern illustrated in Figures 4-7 may The least significant bit pattern can be repeated in parallel as illustrated in FIG. The encoding for the second set of patterns 23 is the third one for the first set 21.
It is the same as shown in the figure. Even if this is done, there will be no difficulty in decoding even at the interface between the two patterns, as shown in the point R in FIG. This can be illustrated with an example. In this way, the intercept is 2
is equal to -8 in complement notation, and the X coordinate is equal to +7. Using equation (2), at point R: D = 1000 or less This result can be confirmed by analyzing the graph in FIG. If necessary, another set of stripe patterns can be added to cover the X direction as desired. What should be noted here is that the Z-axis scale is proportional to the X-sleeve scale, and both scales are dependent on the width of the stripe. The ratio of the Z-axis scale to the X-axis scale depends on the angle θ. The larger the angle θ, the more Z
Axial resolution increases.

Although the invention has been described above with respect to distance measurements of discontinuities in object space, it has great value in creating maps of contoured object surfaces. For example, if the stripe pattern 27 of the present invention is projected onto the sphere 29 shown in FIG.
A curved stripe 31 appears as shown in the figure. Since all pixels along the vertical line of the focal plane 33 of the TV camera have the same X coordinate, the curvature of the stripe 31 above and below the vertical midpoint of the focal plane will move toward the left to any selected vertical direction. This means that the farther the pixels on the line are from the vertical midpoint, the more they are irradiated with higher-numbered stripes. the result,
According to equation (2), an increasing distance measurement is obtained indicating that the surface is curving away from the [11].

Using the present invention, a dense range map of the entire object space can be quickly created.

Many advantages are derived from the disclosed technology. The first advantage is high reliability. This high reliability is due to the fact that there are no moving mechanical elements and the light level is detected binary.
This can be achieved by relaxing the requirements for the /N ratio and by using a simple deciphering method that does not involve any guesswork or ambiguity. Parity bits can be applied to the encoded signal by simply adding one or more exposures with the pattern of parity bits shown in Figure 3. You can check the parity of the history and then move on to decryption. In this way, system reliability is improved by eliminating unreliable pixels caused by unexpected movement of the object or by receiving unwanted secondary reflections from relatively reflective surface properties. I can do it.

Another effect can be obtained by choosing an odd number of parity tests. Darkness for stripe number 0-
The blind spots created by dark-dark coding are eliminated by the bright parity bits created by the odd-numbered parity scheme. According to the present invention, a full range of information can be obtained at high speed.

Our invention requires only frames, not logs.
It's only 2M frames. Here, is the number of resolvable range elements, literally the "dynamic range"
It is. For example, the present invention only requires B frames (i.e. about 174 seconds using a regular video camera), in this case M = 256, and the decoding process does not impose any burden on the overall system. Because,
This is because decoding is a very simple method. For this resolution, a single stripe technique requires as many as 25B frames (over 7 seconds). The reduction in data collection time from this single stripe technique to our technique is (
log 2 M)/M. Roughly speaking, this reduction corresponds to the reduction in calculation time obtained by changing from discrete Fourier exchange to fast Tourier transform. This effect increases the dynamic range requirements.

The output of the optical three-dimensional digital data acquisition system of the present invention is digital in nature. Representing ranges in two's complement numbers facilitates interfacing with a variety of host computers and communication channels in many future applications.

Another advantage of the present invention is that 1, since the system uses a conventional video camera, a conventional gray scale video signal is obtained. As a result, both analog grayscale and digital ranges can be used simultaneously. For example, all points of interest can be labeled with x-y coordinate information superimposed on the same video frame to assist an operator monitoring the robot.

Yet another advantage of the present invention is that the range resolution ΔZ is limited only by the camera resolution ΔX. That is,
ΔZ = (1/lanθ)ΔX, where θ is the oblique angle between the camera and the projector. In this way, magnification can be varied by using different lenses or by adding zoom capability. Therefore, it would be possible to create a dense lens map of telescopic objects, microscopic objects, and intermediate objects between them.

Although specific embodiments of the present invention have been described in detail above,
Those skilled in the art will appreciate that various modifications and substitutions to the embodiments are possible given the general teachings of the disclosure. Accordingly, the specific arrangements disclosed are merely illustrative of the invention and do not limit its scope. The scope of the invention should be determined by the claims appended hereto and their equivalents.

[Brief explanation of the drawing]

FIG. 1 is a perspective view showing the arrangement of an apparatus for carrying out the invention. FIG. 2 is a schematic plan view showing the geometric arrangement of the TV right camera projector of FIG. 1 with respect to an object whose distance to the TV right camera is to be measured. FIG. 3 is a diagram illustrating a system used to encode the drive pattern of light produced by a projector. 4-7 illustrate a continuous light encoded stripe pattern formed by a projector for an exemplary system that divides the pattern into IB-tree stripes. Figure 8 shows the stripe pattern of the least significant bit in the second
It is a diagram showing how the image is projected onto the object shown in the figure and also showing a coordinate system used to calculate the distance. FIG. 9 is a diagram illustrating the method used to determine object distances from data generated by encoded stripe patterns. FIG. 1O is a diagram illustrating a measurement technique according to the invention in which two sets of stripe patterns are projected in parallel. FIG. 11 is a diagram showing an example of application of the present invention to create a map of the surface of a three-dimensional object. l...Projector 3...Space 5...TV right camera...Digital computer 9...Projection axis 11...Optical axis 13.15...Object 17 ... Stripe pattern 0 ... Oblique angle -8-1-@-3-4-5-t -1ul t
Cori co % rFIG, IO 7giFIG, lI

Claims (1)

[Claims] 1. A T having a plurality of pixel positions located in a place where light can be detected and arranged on the X-Y plane and an optical axis perpendicular to the X-Y plane.
A method for measuring distances between multiple points on an object arranged in a three-dimensional object space using a V-camera and a projector, the method comprising: measuring a continuous pattern of uniquely encoded light stripes; projecting onto the object along an axis oblique to the optical axis of the TV camera such that the object reflects a portion of the coded light towards the TV camera; wherein the stripes are encoded by selectively brightening and darkening each stripe in a unique sequence of said continuous pattern; and a detected reflection at each pixel location for said continuous pattern of light stripes. recording a coded pattern of light, and determining the distance from the TV camera of a point that reflects light toward each pixel using the coded light signal detected by the X-Y coordinates of the pixel and the pixel position; and calculating as a function of . 2. The method of claim 1, wherein the number of consecutive patterns is equal to the logarithm of the number of stripes in each pattern, with 2 being a constant. 3. The method of claim 2, wherein the pattern of stripes is repeated in parallel. 4. projecting another pattern of encoded stripes to generate a specific parity bit for each stripe; calculating the parity value of the encoded optical signal sensed at each pixel; 3. The method of claim 2 further comprising the step of erasing the distance calculated for the encoded optical signal sensed at any pixel that does not have a parity value. 5. A T having multiple pixel positions arranged on the X-Y plane and an optical axis perpendicular to the X-Y plane, located in a place where light can be detected.
A method of measuring distances between multiple points on an object arranged in a three-dimensional object space using a V-camera and a projector, in which the X-axis and Y-axis are in the pixel array of the TV camera. creating an X-Y-Z coordinate in the object space parallel to the corresponding axis and with the Z axis parallel to the optical axis of the camera; and encoded stripes of light parallel to the Y axis of the coordinate. is projected into the object space and onto the object along an axis that intersects the X and Z axes at an oblique angle so that the object directs a portion of the light toward the TV camera. setting a scale on each of the X and Z axes of the coordinates and the diagonal width of the stripe, and as a function of the scale set in the modulus of the X axis; from the origin of the X-Y-Z coordinates by setting a scale on the Encoding the stripes with consecutive numerical numbers across each pattern and Z of points that reflect light towards each pixel.
calculating an axial distance by subtracting the x-coordinate of the pixel from the encoded signal sensed at the pixel location. 6. The method of claim 2, wherein the pattern of stripes is repeated in parallel. 7. The origin of the X-Y-Z coordinates is placed approximately at the center of the target space, and the stripes are arranged in a direction opposite to the positive number sequence arranged in one direction from the origin along the Y-axis. 5. The method of claim 4, wherein the stripe passing through the origin is encoded as a negative number sequence and the stripe passing through the origin is encoded with the value 0. 8. Claim 7, wherein the pattern of light stripes is spatially repeated in parallel, and corresponding stripes in each repetition have the same encoding.
Section method. 9. projecting another pattern of encoded stripes to generate a specific parity bit for each stripe; calculating a parity value of the encoded optical signal sensed at each pixel; 3. The method of claim 2 further comprising the step of erasing the distance calculated for the encoded optical signal sensed at any pixel that does not have a parity value. 10. The method of claim 7, wherein the stripes encoded as negative numbers are encoded as two's complement notation. 11. Projecting another pattern of the stripes encoded such that each stripe is an odd number; calculating a parity value of each detected encoded optical signal at each pixel position; and detecting the encoded optical signal. Claim 1, further comprising the step of erasing the distance calculated for any pixel for which the optical signal is not an odd number.
The method described in item 0. 12. A device for determining the Z-axis position of a point on an object surface in an X-Y-Z coordinate system, in which a plurality of pixels irradiated with light are A TV camera arranged in a Y array, with its optical axis parallel to the Z axis, and equipped with means for generating an X coordinate signal in binary for each pixel position; A continuous pattern of encoded light stripes is projected onto the object along an axis that intersects the optical axis of the TV camera at an oblique angle so that the object directs the coded light toward the TV camera. A projector for reflecting a part of the
The scales of the X-axis and Z-axis of the coordinate system are defined by the diagonal width of the stripe, and the two axes and the X-axis scale of the pixel array are defined by the scale of the X-axis of the coordinate system; means for uniquely encoding each stripe with consecutive numerical numbers across each pattern by selectively brightening and darkening each stripe in said continuous pattern of a uniqueness sequence expressed in binary; means for calculating the Z-axis position of a point on the object surface by subtracting an X-coordinate signal for each pixel of the TV camera from the encoded reflected light signal received by the pixel; A device characterized by: 13. The projector comprises means for spatially repeating the continuous pattern of stripes of light, and the encoding means encodes corresponding stripes of each repetition to the same code. 13. Apparatus according to claim 12, characterized in that it comprises means for. 14. The optical axis of the TV camera is aligned so that it is coaxial with the Z axis of the coordinate system, and the projector is arranged such that the stripe pattern crosses the X axis of the coordinate system and the stripe pattern crosses the X axis of the coordinate system. projecting the pattern so that one part is on one side of the coordinate origin, the other part is on the other side, and one line passes through the origin; 13. The apparatus according to claim 12, wherein the same part is encoded as a continuous positive binary number, the same part is encoded as a continuous negative binary number, and one stripe passing through the origin is encoded as a zero. 15. The encoding means encodes negative 2 using 2's complement notation.
15. Apparatus according to claim 14, characterized in that the other parts of said stripes are encoded in base numbers. 16. said projector projecting another pattern of stripes of light; said encoding means encoding said other pattern into an odd number of parity bits for each stripe; and said computing means 16. The apparatus of claim 15, wherein the apparatus also calculates the parity of the encoded optical signal received by each pixel and rejects odd non-parity stripes. 17. The projector comprises means for spatially repeating the continuous pattern of stripes of light as described above in parallel, and further that the encoding means encodes corresponding stripes of each repetition with the same code. 15. Apparatus according to claim 14, characterized in that it comprises means for digitizing. 18. A device for determining the position of a point on the surface of an object, including a TV camera in which a plurality of pixels receiving light are arranged in an array on the X-Y plane and the optical axis extends toward the object. , projecting a continuous pattern of stripes of light onto the object along an axis that intersects the optical axis of the TV camera at an oblique angle so that a point on the object surface directs the encoded light toward the TV camera. a projector for reflecting a portion of said TV; means for uniquely encoding said stripe by selectively brightening and darkening each stripe in a unique sequence of said continuous pattern of stripes; The position of the point on the object surface in the direction parallel to the optical axis of the camera is determined by the X-Y coordinates of each pixel that receives light reflected from these points and the encoded value of the reflected light signal. and means for calculating the Y coordinate as a function of the encoded value of the reflected light signal. 19. The projector comprises means for spatially repeating the continuous pattern of stripes of light as described above in parallel, and further that the encoding means encodes corresponding stripes of each repetition with the same code. 19. Apparatus according to claim 18, characterized in that it comprises means for digitizing.